Methods, systems, and computer program products for signature detection. One example of a method includes: acquiring an article defect density map comprising a plurality of sections corresponding to a first resolution level which is indicative of defect numbers for the sections, and determining a distribution representative of the defect numbers or function thereof; determining a threshold in accordance with said distribution, and identifying sections, out of said plurality of sections in the article defect density map, with defect numbers or function thereof above the threshold; and clustering at least part of adjoining identified sections, into one or more signatures, thus detecting said one or more signatures.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A system for signature detection, comprising: a memory; and a processor, coupled to the memory, to: acquire an article defect density map comprising a plurality of sections corresponding to a first resolution level which is indicative of defect numbers for the sections, and to determine a distribution representative of the defect numbers or function thereof; determine a threshold in accordance with said distribution, and to identify sections, out of said plurality of sections in the article defect density map, with defect numbers or function thereof above the threshold; and cluster at least part of adjoining identified sections, into one or more signatures, and thus configured to detect said one or more signatures.
A system detects signatures (patterns of defects) on an article. The system uses a processor and memory. The processor creates an article defect density map, dividing the article into sections. Each section's defect count is determined. A distribution of defect counts across all sections is calculated. A threshold is determined based on this distribution. Sections with defect counts above the threshold are identified and clustered together if they are adjoining. These clusters are the detected signatures.
2. The system of claim 1 , wherein to acquire said article defect density map, the processor is to acquire an article defect map, to divide at least part of said article defect map into the plurality of sections corresponding to the first resolution level, and to count defects in each of the sections, thereby giving rise to the defect numbers for the sections.
The system from the previous description creates the article defect density map by first acquiring a raw article defect map. This raw map is then divided into sections. The number of defects in each section is counted to determine the defect numbers for each section, thus creating the defect density map.
3. The system of claim 1 , wherein to acquire said article defect density map, the processor is to acquire an article defect map, to divide at least part of said article defect map into the plurality of sections corresponding to the first resolution level, to count defects in each of the sections, to acquire a scanned portion map, and for at least one of the sections to use the scanned portion map to normalize the count by one or more of: scanned portion of the section or size of the section, thereby giving rise to the defect numbers for the sections.
The system from the previous description creates the article defect density map by acquiring a raw article defect map and dividing it into sections, counting defects in each section. A scanned portion map is also acquired. For each section, the defect count is normalized, using the scanned portion map, by considering the scanned portion of the section or the size of the section. This normalization accounts for uneven scanning and ensures accurate defect density.
4. The system of claim 1 , wherein the plurality of sections is defined by polar coordinates.
In the system from the initial signature detection description, the sections used to create the article defect density map are defined using polar coordinates (radius and angle) instead of Cartesian coordinates (x, y). This is useful for articles with circular or curved shapes.
5. The system of claim 1 , wherein for each of one or more other resolution levels, the processor is to acquire the article defect density map comprising a plurality of sections corresponding to the other resolution level, to determine a distribution, to determine a threshold, to identify sections, to cluster into one or more signatures to detect said one or more signatures for the other resolution level, and to unify at least part of overlapping detected signatures from a plurality of resolution levels.
The system from the initial signature detection description also performs signature detection at multiple resolution levels. For each resolution level, it acquires an article defect density map, determines a distribution and threshold, identifies sections with high defect counts, and clusters them into signatures. Signatures detected at different resolution levels are then unified, merging overlapping signatures to provide a comprehensive view of the defect patterns.
6. The system of claim 1 , wherein the processor is to detect at least one polar signature using at least one article defect density map, each comprising a plurality of sections defined by polar coordinates corresponding to a respective resolution level, and to detect at least one Cartesian signature using at least one article defect density map, each comprising a plurality of sections defined by Cartesian coordinates corresponding to a respective resolution level, and to combine at least part of at least one detected Cartesian signature and at least part of at least one detected polar signature into a combined detected signature.
In the system from the initial signature detection description, the system detects signatures using both polar coordinates and Cartesian coordinates. It detects polar signatures using defect density maps with sections defined by polar coordinates and Cartesian signatures using defect density maps with sections defined by Cartesian coordinates. The system then combines the detected Cartesian and polar signatures into a single combined signature for better coverage of different defect types.
7. The system of claim 6 , wherein for a detected Cartesian signature and a detected polar signature with at least one defect in common, the processor is to determine whether or not to combine at least part of said detected signature and at least part of said polar signature, depending on a result of at least one test, wherein said at least one test includes one or more of: a radius test indicative of a relationship between an average radius for defects in the detected Cartesian signature and inner and outer radial limits of the detected polar signature, or a density test indicative of a relationship between a density of the detected Cartesian signature and a density of the detected polar signature.
In the system combining Cartesian and polar signatures as described above, when a Cartesian signature and a polar signature share at least one defect, the system determines whether to combine them based on tests. These tests include a radius test (comparing the average radius of defects in the Cartesian signature to the radial limits of the polar signature) and/or a density test (comparing the defect density of the Cartesian signature to the defect density of the polar signature).
8. The system of claim 1 , wherein each at least one of the one or more detected signatures, is graded with a grade that is at least partly dependent on the threshold.
In the system from the initial signature detection description, each detected signature is assigned a grade. This grade is partially dependent on the threshold used to identify the sections that form the signature. Signatures formed from sections exceeding a higher threshold may receive a higher grade.
9. The system of claim 1 , wherein the processor is further to perform, for a detected signature which includes at least part of at least one of said one or more detected signatures, any of a group of actions, said group including: characterizing, filtering, or separating at least partly by at least one of: radius, angle, density, area, grade, number of defects, ratio of number of defects, resolution gap, detection procedure, type, number of sections, defect locations, superclusters, interest real defects, nuisances, or false alarms.
The system from the initial signature detection description further processes detected signatures by characterizing, filtering, or separating them based on various attributes. These attributes include radius, angle, density, area, grade, number of defects, ratio of defects, resolution gap, detection procedure, signature type, number of sections, defect locations, superclusters, whether they represent genuine defects, nuisances, or false alarms.
10. The system of claim 1 , wherein said article density map is a density map for an article which is at least one of: round, or has a perimeter that is at least partly curved.
In the system from the initial signature detection description, the article defect density map is created for an article that is either round or has a perimeter that is at least partly curved.
11. A method of signature detection, comprising: acquiring an article defect density map comprising a plurality of sections corresponding to a first resolution level which is indicative of defect numbers for the sections, and determining a distribution representative of the defect numbers or function thereof; determining a threshold in accordance with said distribution, and identifying sections, out of said plurality of sections in the article defect density map, with defect numbers or function thereof above the threshold; and clustering at least part of adjoining identified sections, into one or more signatures, thus detecting said one or more signatures.
A method detects signatures (patterns of defects) on an article. The method involves creating an article defect density map, dividing the article into sections. Each section's defect count is determined. A distribution of defect counts across all sections is calculated. A threshold is determined based on this distribution. Sections with defect counts above the threshold are identified and clustered together if they are adjoining. These clusters are the detected signatures.
12. The method of claim 11 , wherein said acquiring said article defect density map includes: acquiring an article defect map, dividing at least part of said article defect map into the plurality of sections corresponding to the first resolution level, and counting defects in each of the sections, thereby giving rise to the defect numbers for the sections.
The method from the previous description creates the article defect density map by first acquiring a raw article defect map. This raw map is then divided into sections. The number of defects in each section is counted to determine the defect numbers for each section, thus creating the defect density map.
13. The method of claim 11 , wherein said acquiring said article defect density map includes: acquiring an article defect map, dividing at least part of said article defect map into the plurality of sections corresponding to the first resolution level, counting defects in each of the sections, acquiring a scanned portion map, and for at least one of the sections using the scanned portion map to normalize the count by one or more of: scanned portion of the section, or size of the section, thereby giving rise to the defect numbers for the sections.
The method from the previous description creates the article defect density map by acquiring a raw article defect map and dividing it into sections, counting defects in each section. A scanned portion map is also acquired. For each section, the defect count is normalized, using the scanned portion map, by considering the scanned portion of the section or the size of the section. This normalization accounts for uneven scanning and ensures accurate defect density.
14. The method of claim 11 , wherein the plurality of sections is defined by polar coordinates.
In the method from the initial signature detection description, the sections used to create the article defect density map are defined using polar coordinates (radius and angle) instead of Cartesian coordinates (x, y). This is useful for articles with circular or curved shapes.
15. The method of claim 11 , wherein said method comprises for each of one or more other resolution levels: acquiring an article defect density map comprising a plurality of sections corresponding to the other resolution level, determining a distribution, determining a threshold, identifying sections, and clustering into one or more signatures, thus detecting said one or more signatures for the other resolution level, said method further comprising: unifying at least part of overlapping detected signatures from a plurality of resolution levels.
The method from the initial signature detection description also performs signature detection at multiple resolution levels. For each resolution level, it acquires an article defect density map, determines a distribution and threshold, identifies sections with high defect counts, and clusters them into signatures. Signatures detected at different resolution levels are then unified, merging overlapping signatures to provide a comprehensive view of the defect patterns.
16. The method of claim 11 , wherein said method comprises: detecting at least one polar signature using at least one article defect density map, each comprising a plurality of sections defined by polar coordinates corresponding to a respective resolution level, and detecting at least one Cartesian signature using at least one article defect density map, each comprising a plurality of sections defined by Cartesian coordinates corresponding to a respective resolution level, said method further comprising: combining at least part of at least one detected Cartesian signature and at least part of at least one detected polar signature into a combined detected signature.
In the method from the initial signature detection description, the method detects signatures using both polar coordinates and Cartesian coordinates. It detects polar signatures using defect density maps with sections defined by polar coordinates and Cartesian signatures using defect density maps with sections defined by Cartesian coordinates. The method then combines the detected Cartesian and polar signatures into a single combined signature for better coverage of different defect types.
17. The method of claim 11 , wherein each at least one of the one or more detected signatures, is graded with a grade that is at least partly dependent on the threshold.
In the method from the initial signature detection description, each detected signature is assigned a grade. This grade is partially dependent on the threshold used to identify the sections that form the signature. Signatures formed from sections exceeding a higher threshold may receive a higher grade.
18. The method of claim 11 , wherein said method further comprises: performing for a detected signature which includes at least part of at least one of said one or more detected signatures, any of a group of actions, said group including: characterizing, filtering, or separating at least partly by at least one of: radius, angle, density, area, grade, number of defects, ratio of number of defects, resolution gap, detection procedure, type, number of sections, defect locations, superclusters, interest, real defects, nuisances, or false alarms.
The method from the initial signature detection description further processes detected signatures by characterizing, filtering, or separating them based on various attributes. These attributes include radius, angle, density, area, grade, number of defects, ratio of defects, resolution gap, detection procedure, signature type, number of sections, defect locations, superclusters, whether they represent genuine defects, nuisances, or false alarms.
19. A computer program product comprising a non-transitory computer useable medium having computer readable code embodied therein for signature detection, the computer program product comprising: computer readable program code for causing a computer to acquire an article defect density map comprising a plurality of sections corresponding to a first resolution level which is indicative of defect numbers for the sections, and to determine a distribution representative of the defect numbers or function thereof; computer readable program code for causing a computer to determine a threshold in accordance with said distribution, and to identify sections, out of said plurality of sections in the article defect density map, with defect numbers or function thereof above the threshold; and computer readable program code for causing a computer to cluster at least part of adjoining identified sections, into one or more signatures, and thus to detect said one or more signatures.
A computer program product stored on a non-transitory medium detects signatures (patterns of defects) on an article. The program code causes a computer to create an article defect density map, dividing the article into sections. Each section's defect count is determined. A distribution of defect counts across all sections is calculated. A threshold is determined based on this distribution. Sections with defect counts above the threshold are identified and clustered together if they are adjoining. These clusters are the detected signatures.
20. The computer program product of claim 19 , wherein the plurality of sections is defined by polar coordinates.
In the computer program product from the previous description, the sections used to create the article defect density map are defined using polar coordinates (radius and angle) instead of Cartesian coordinates (x, y). This is useful for articles with circular or curved shapes.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
March 30, 2015
April 4, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.